The term "flipped" has caused much confusion. Collo explained here how "flipping" (rotating a QRD panel 180 deg.) is not really going to do much
-- instead consider an inverse or divergent prime number panel.
"Because the panels are basically symmetrical, your "flipped" [180 deg.] panel would give the same diffusion pattern - no use as the alternate panel in a Barker sequence." Look at the pics so see the difference between "flipped" as in rotated, and a proper inverse.
If you visit the QRDUDE software
(free QRD design software by Collo) guide page here
,and scroll down to the "Multiple panels"
section, you find that as well as a proper inverse, panel modulation can be done. Scroll down just a bit more to the line "There are several possibilities for the modulating panel to be used:
" with the examples and pics. See: Option #4
: "Rotate a standard panel by 180 degrees. This is the simplest solution, although the change in diffusion angles is small compared to that obtained with the other approaches. The panel is equivalent to a standard panel that has been left-shifted one well..."
Note that: "The modulation works best at multiples of the design frequency. Only at these frequencies do the diffuser and its inverse create exactly opposite pressures. At other frequencies, the modulation is likely to help with the scattering as it breaks up periodicity lobes, but in a more uncontrolled manner." (p. 251, Acoustic Absorbers and Diffusers: Theory, Design and Application
By Trevor Cox, Peter D'Antonio). Trevor replied here & here to a letter I wrote him
on the question of sticking to the design frequency for modulation. Here is the Q&A:
Me: My question is a simple one. My N23 design freq. = 994Hz, Max Well Depth [MWD] = 135mm. With this arrangement: N23 N23 N19 N23 (Barker code: +1 +1 -1 +1) To make N19 design freq. 994Hz (max well depth 127mm) or disregard the design Hz, and just build to the same depth (135mm) as the N23?
I ask, in that from reading it seems that matching the design Hz is preferable -- though some on the board disagree, stating that a mismatch is actually preferable, in terms of modulation. Could you clarify this point?
Trevor replies: Probably impossible to answer the question without a few predictions. But given you are using relatively large N primes, flat plate frequencies are not a worry, so choosing the same design frequency is OK (choosing a different one is probably OK as well!)
Due to the above, I altered my array for 4 N23 panels to this (matching MWD in the two different panels, and selecting iN23 instead of N19 for modulation):
Barker code +1 = N23+0
: design freq. = 994Hz = 135mm MWD -1 = iN23+4
: design freq. = 1055Hz = 135mm MWD iN23(+4)
is an inverse N23, shifted up 4 depth units (and left 9 wells), which allows maximum depth reduction (as possible) for the deepest well. The left shift allows the frame sides to be used as part of the furthest left- and right-side wells. I used QRDUDE
to run my calculations.
Here is a picture of the array. The 2 rows of four panels measure 4 meters wide by 2 meters high on a 5 meter wide wall (1 panel's height = 94cm):
Absorption hasn't been installed yet. You can see the iN23+4 panels as they have the spacers in the middle of 4 consecutive deep wells (a couple spacers need yet to be affixed). Collo wrote
: "The iN23+4 is not the inverse of the N23+0 -- Having said that, you are maximising the "free ride" for both, thus getting the lowest design frequencies for each. Given Trevor's earlier comments downplaying the importance of having an exact inverse, this is should be as good as any other non-exact arrangement."
My conclusion, "This seems useful, in that now there are two decent modulation options beyond the iN23+0. I can use an N19, or the iN23+4. In terms of studio decor, the iN23+4 will look much better and cover a larger area." I hope this info is helpful.
If you wish to experience pain, read the thread of my own QRD thrashing (don't do it!) from here
. Some pics of how the build went with XPS foam core construction are here
. The term "flipping" as a primary solution for QRD modulation needs further discussion. I highly
recommend QRDUDE and the accompanying explanatory pages -- a real lifesaver. Much thought and expertise has gone into its creation. My live room has benefited in a lively way from the N23/iN23 array.
PS These panels cost about $50. per in materials, including stain and finishing, in Japan (where all materials are more expensive than in N.Amer.). The top/bottom well inserts of hard rubber (the black areas) added about $8.00. So, call it $60. per panel. It did take some time.
Originally Posted by Dange
From Cox and D'Antonio, the ones flipped according to the barker sequence perform better than a single period (one QRD) and the repeated period (identical repeated QRD)....